• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 64 subscribers
  • Views 183 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPS281C100: Device failure during short to ground on output

Cezar Chirila
Prodigy 30 points
Part Number: TPS281C100

Tool/software:

Hi, 

I am using a TPS281C100 device to switch a resistive load. I have built a small test board and during testing I have found that during a short-to-ground test on the output (so VOUT of TPS281C100A was connected to ground before EN was high, then EN pin has been switched high while maintaining the short to ground) has concluded in failure of the switch (loud pop and can visibly see damage on the chip). The device does not work anymore after this event. 

Before testing the short-to-ground condition, the device was working as expected. 

I have attached a schematic below. Have I used the device wrong or have I misunderstood how the shortcircuit protection works?

For reference, at the moment of testing, VS was 60V. 

  • 0
    TI__Mastermind 32175 points

    Hello Cezar,

    Apologies for the delay in answering. Could you share on the exact loading method you are shorting the device to ground? During validation we did do bench testing for a 60V short-to-ground event and the device survived. You wouldn't happen to have caught a wave form also, would you?

    Also- if rerunning the test,  would it be possible to change the input capacitance to something like 1uF for the sake of testing?

    Best Regards,
    Tim 

  • 0 in reply to Timothy Logan
    Prodigy 30 points

    Hi Timothy, 

    Unfortunately I did not capture the waveform. The output was shorted to ground after applying power, but before enabling the output via the 'EN' pin. The short was created using a simple wire (so very low impedance to ground).

    On rerunning the test on the devices that survived, I cannot replicate the problem. (Even without adding the 1uF input capacitance)
    Would you mind explaining how would the extra input capacitance help? Would it prevent getting the device in an unknown state due to brown out?

    My only worry is that there is something marginal in my design that will cause a certain proportion of devices to fail in the field. If you think this is a one time event, maybe due to bad assembly or user testing error I am happy to close the problem. 

    One last detail I forgot to mention, which might be important is that the device is mounted on a two sided flex PCB with an area of about 2cm^2, where the bottom side is full copper (1oz) and the device is connected to this plane using vias, as suggested in the datasheet in the 'LAND PATTERN EXAMPLE'. I am mentioning this because I appreciate that the area is fairly small, and I wonder if this could have an impact on device performance? On the devices that did not fail, the thermal limit was reached fairly soon.  

    Thank you!

    Sincerely, 

    Cezar

  • 0 in reply to Cezar Chirila
    TI__Mastermind 32175 points

    Cezar,

    For short-to-ground, the biggest factor here would be peak power. The device reacts to the short-to-ground event within a matter of ~5uS or so (by engaging the current limit), so the level of impedance on the "short" would play a factor here. For automotive applications, for example, the standard specification is 5uH and 50mOhm- however for industrial it tends to be more defined by the customer requirements. We do test the device for 60V shorts as a use case- so I would not say there is anything marginal on your design. 

    The capacitor on the input would help smooth out the peak power pulse when the short is applied. For a "hot short" (output is suddenly shorted), the capacitances helps any oscillation on the input/output. 

    Thanks for they layout information- short conditions tend to happen very quickly so the layout itself does not have too much of an impact. Most of the heat dissipation is handled by the package itself so no issues with your layout parameters.

    Best Regards,
    Tim